Title :
On convergence of fuzzy reinforcement learning
Author :
Berenji, Hamid R. ; Vengerov, David
Author_Institution :
Computational Sci. Div., Intelligent Inference Syst. Corp., Mountain View, CA, USA
Abstract :
This paper provides the first convergence proof for fuzzy reinforcement learning. We extend the work of Konda and Tsitsiklis (2000), who presented a convergent actor-critic algorithm for a general parameterized actor. In our work we prove that a fuzzy rule base actor satisfies the necessary conditions that guarantee the convergence of its parameters to a local optimum. Our fuzzy rule base uses the Takagi-Sugeno-Kang rules, Gaussian membership functions and product inference.
Keywords :
Markov processes; convergence; fuzzy logic; fuzzy set theory; inference mechanisms; learning (artificial intelligence); Gaussian membership functions; Markov decision process; Takagi-Sugeno-Kang rules; actor-critic algorithm; convergence; fuzzy reinforcement learning; fuzzy rule base actor; necessary conditions; parameterized actor; product inference; Collaborative work; Computational intelligence; Convergence; Function approximation; Fuzzy set theory; Inference algorithms; Intelligent agent; Intelligent systems; Learning; NASA;
Conference_Titel :
Fuzzy Systems, 2001. The 10th IEEE International Conference on
Print_ISBN :
0-7803-7293-X
DOI :
10.1109/FUZZ.2001.1009030
